This project will examine how the physical, social, and economic environments influence dietary choices and, thereby, contribute to disparities in obesity rates. The present hypothesis is that the low cost and easy access to energy-dense foods may be an independent predictor of higher obesity rates. This study will combine survey research techniques with novel methods of spatial analysis to obtain in-depth information about the participants'physical and food environment and their access to food sources. A telephone survey of a stratified sample of 2,000 adult residents of King County, WA, will provide self-reported data on the participants'socioeconomic position (SEP), shopping and eating habits, food expenditures, obesity, and health. All addresses will be geocoded to facilitate spatial analysis and mapping. Objective measures of SEP will be based on data at the most detailed resolution available -- the County tax assessor at the tax parcel level. Objective measures of physical access to three key food sources, defined for each respondent in terms of quality, distance and time, will be calculated at the individual level using custom GIS based applications and novel metrics of the physical environment. Custom programs within GIS software will capture and perform preliminary geometric analysis of geospatial data within the proximity of the survey participants'home environment and their chief food sources. Dietary energy density and monetary cost of individual diets will be estimated using local food prices and an additional diet questionnaire instrument that will be mailed to all 2000 study respondents. Local variation in food prices will be assessed using a market basket technique. Spatial analyses and multilevel regression models will be used to test the hypothesis that physical and economic access to foods sources may predict dietary energy density and obesity rates.